The Voting-Machine Suit Against Diebold Deserves to Be Heard
Before I excerpt, two points:
- Nobody can credibly claim that Diebold or e-voting had anything to do with the 2000 presidential election dispute in Florida. E-voting wasn’t used anywhere at that point.
- Nobody can credibly claim that Diebold or e-voting had anything to do with the flawed belief that Ohio somehow was really won by John Kerry. I don’t think e-voting was in place anywhere in Ohio at the time. The central argument of the Kerry-won-Ohio holdouts, which I believe is bogus, is that people were prevented from voting, not that the votes weren’t tabulated properly.
So the lawsuit covered here is about future elections, not previous ones, despite plaintiff Robert F. Kennedy Jr.’s and In These Times reporter John Ireland’s carping to the contrary.
And despite his strained arguments on previous elections, and the strained reporting contained in most of the linked article, Kennedy’s argument against Diebold needs to be heard:
….. (Kennedy claims that) Diebold and other electronic voting machine (EVM) companies fraudulently represented to state election boards and the federal government that their products were “unhackable.â€
Kennedy claims to have witnesses “centrally located, deep within the corporations,†who will confirm that company officials withheld their knowledge of problems with accuracy, reliability and security of EVMs in order to procure government contracts.
Unfortunately, there is more than a little evidence (here, here [third item at link], and here [final item at link]) that e-voting machines are anything but hackproof.
And if they’re not, they shouldn’t be used. Period. This shouldn’t even resemble being a partisan issue.
While I agree that there’s no reason to dismiss the lawsuit, I do disagree with your assertion that electronic voting must be unhackable. EVERY operating computer system is hackable. Are paper ballots perfect? If they were, no one would even consider e-voting. The perfect is the enemy of the good, and in this case the standard must include the totality of the situation, i.e. if I have to hook a laptop up to the voting machine to hack it, or cut off a padlock, then those are not reasonable threats since the poll monitors won’t let me do that. Bottom line, I don’t know how hackable Diebold’s equipment is (but given that an awful lot of banks trust their ATMs, I’m guessing “not very”), but that is the real issue. Expecting perfection is unrealistic.
Comment by Steve — July 18, 2006 @ 10:09 am
Absolutely. This is not partisan at all. There are three main prolems that I have with EVM’s, all technical. First, as you mentioned, they are vulnerable to security exploits. The only online systems that are not vulnerable, are those with the network connection disconnected :). Free elections are the pinnacle of democracy. It would be a travesty if elections results were tampered with by a cracker. My second problem is reliability of the systems. PC’s, servers, and networking equipment, of every class and grade fail at some point. Corporations design systems with this fact in mind. We build fault tolerance and high availability into our infrastructure. Voting machines should be no different. If we use EVM’s, we should absolutely have corresponding paper ballots printed for each vote. Finally, closed source code. Software source code is heavily guarded. No company will allow access to this code, which introduces transparency problems. A rogue developer could write a random algorithm to subtract votes from or add votes to a candidate or party of their choosing. While an internal corporate code audit could discover this, it would not be in the company’s best interest to announce the flaw. It would most likely be a discovered “bug” that the company fixes quietly.
Comment by Kevin irwin — July 18, 2006 @ 12:12 pm
#1, My definition of “hackproof” is really “very, very, VERY difficult to hack, and if somehow successfull hacked, detected IMMEDIATELY.
#2, You get close to a real important point. With EVMs, you really need to have a tech specialist on site (or maybe floating between 3-4 close precincts) to troubleshoot problems, because voter should not be inconvenienced once they get to the polls.
SOMEBODY better get to see the code (outside CPA firms, for example), or we won’t know if algorithms are ALREADY in there to mess with the voting. And yes, we need a corresponding paper trail (a receipt to the voter and a live-updated printout/file of each vote for the Board of Elections). I don’t think Diebold is going that route.
Comment by TBlumer — July 18, 2006 @ 3:19 pm